Change

Server Virtualization is considerably new innovation that was implemented by my organization and my team, to increase application performance, achieves better utilization of hardware infrastructure and cost reduction for power and cooling.  The implementation of this technology helped us to reduce server-provisioning time by 75%, and made it easier for the system administrator to manage the huge number of servers we have.

The change will be in migrating physical systems into virtualized ones managed by one interface, and implement additional features that come along with this technology such as High availability and live migration of virtualized servers incase of hardware failure. The change is still in implementation phase, and yet to complete by end of august. CTO, CIO and COO supported this technology as they fully understand and realize its importance and the benefits behind it. We came with this proposal based on a requirement to reduce data center operation cost, licenses cost and personal cost. In addition to that, risk and business continuity department were looking for a fast, cost effective and not risky solution to build Disaster Recovery Site, and to make all business critical and business important application available in DR Site, all those supporters from top management and risk department made the approval process faster.

To achieve a successful implementation, we engaged application vendors to assist us in the migration process and verify that their applications are supported under the virtualized infrastructure. In this stage we faced a lot of resistance from application vendors, the main claim was because their applications were not tested and its functionality were not verified under this environment. For us we were not convinced with this reason because all operating systems we use are supported, in addition to the fact that all development and test systems for those applications already run under virtualized systems, but this showed lack of proper understanding of this technology from their side. However, as we cannot go forward without their official support, we had to create pilot systems for each application that are identical to production systems, to give them a sort of comfort and an environment to check application functionality, performance and reliability. This was an easy to achieve by costly solution, because the hardware infrastructure was available, we didn’t have to purchase new hardware for the pilot implementation. But this took lots of effort from our side to keep duplicate systems and isolate their network completely to avoid data inconsistency.

This approach helped both of us, as it gave them confidence and made them support the application under virtualized environment. At the end we benefited and the ROI was really high, as we able to reduce power consumption in our data center, easier manageability for system administrators, faster provisioning for new servers, systems resilience, and higher utilization and performance for application servers.

Reference:

Hassell, J., (2007), 'Server Virtualization: GETTING STARTED', Computerworld, 41, 22, p. 31, Computers & Applied Sciences Complete, EBSCOhost, [Online]. (Accessed 25 June 2011)

McAllister, N., (2007), 'Server Virtualization. (cover story)', InfoWorld, 29, 7, pp. 20-22, Computers & Applied Sciences Complete, EBSCOhost, [Online]. (Accessed 25 June 2011)